Valimissüsteemi muudatuste hindamine Eesti parlamendivalimiste näitel

Abstract

The aim of this research was to assess seven theoretical changes of Estonian electoral system based on four evaluation criteria: simplicity, extent of the change, proportionality and fragmentation. The seven changes explored were: replacement of modified d’Hondt with regular or the sequence of Sainte-Laguë, applying natural quota or Droop quota and largest remainders method, increasing district magnitude and losing or decreasing national threshold. To give an assessment to these changes of electoral rules, an experiment was carried out with data from the last two elections of Estonian parliament in 2007 and 2011. Effects of the change were considered positive if they made the electoral system simpler, did not assume huge changes in the existing system, gave more proportional results and did not increase fragmentation considerably. Based on the criteria and the data, the best changes were replacing modified d’Hondt with regular one in the compensational tier and decreasing the threshold to 3% or losing it altogether. Both of these changes made the electoral system less complicated, they were not large-scale changes, they brought about a more proportional seat distribution and they did not make the parliament considerably more fragmented than the actual electoral system. It is important to note that the results of this research have their limits. The experiment showed that from the perspective of proportionality and fragmentation, same changes of electoral system may give different results. The effects of the particular system do not depend only on electoral rules, but importantly also on the input – distribution of electorate’s votes among parties. The input of the two elections, whose data was used in the experiment, was quite different. That made possible that the same changes affected proportionality and fragmentation differently in each case. To make comprehensive conclusions, it is essential to increase the number of cases studied and test the results on broader data. Secondly, as more or less every aspect of the party system is affected by the electoral system, it is important to use more criteria than in this work in assessing the effects of changing the electoral system.